Multi-angle meta-analysis of the gut microbiome in Autism Spectrum Disorder: a step toward understanding patient subgroups Original paper

What was studied?

This ASD gut microbiome meta-analysis examined whether reproducible microbial signatures exist in Autism Spectrum Disorder when raw sequencing data from multiple studies are reprocessed through a single standardized pipeline. The paper integrated 13 datasets representing 10 cohorts, including 16S rRNA gene amplicon, metagenomic, and PhyloChip data, to test how age, sex, and bowel function influence observed microbial differences between ASD and neurotypical children. The authors screened 52 metadata variables and evaluated microbial associations at taxonomic levels ranging from phylum to strain. Their central goal was to disentangle ASD-related microbiome signals from confounding factors and identify microbial taxa that remain significantly associated with ASD across datasets.

Who was studied?

The study encompassed 690 children aged 2–18 years across the United States, Canada, China, Italy, and India. Case–control cohorts varied in size (ASD n=20) and included both sexes, though males predominated, mirroring ASD’s higher prevalence in boys. Importantly, only a minority of prior datasets reported stool consistency, and many had wide age ranges without consistent statistical adjustment. By harmonizing raw data and metadata across cohorts, the authors established comparable subject subsets, particularly focusing on participants with normal bowel function, to more clearly detect ASD-related microbial features.

Most important findings

The analysis revealed that age, sex, and bowel function are major confounders that substantially alter ASD-microbiome associations. Community-level differences (beta diversity) attributed to ASD often disappeared once bowel dysfunction was accounted for. Several taxa previously reported as ASD-enriched or depleted were shown to reflect constipation or diarrhea rather than ASD itself. When restricting analyses to children with normal bowel function, reproducible ASD-associated taxa emerged. Key ASD-depleted organisms included Bacteroides stercoris, Granulicatella elegans, and Massilioclostridium coli. The strain Clostridium M bolteae was consistently enriched in ASD, aligning with its history as an ASD-associated species and its known immunogenic polysaccharide capsule. Strain-level signals were clearer than higher-level taxonomic summaries, supporting the value of fine-resolution microbiome profiling. Additionally, age-stratified analyses showed that certain taxa differed only in childhood or adolescence, underscoring microbiome maturation effects. Sex-specific analyses revealed that ASD-associated taxa in males and females barely overlapped, indicating sex-dependent microbial signatures within ASD cohorts.

Key implications

The work demonstrates that many previously reported ASD-microbiome associations were confounded by unreported variables, especially bowel dysfunction. By controlling for these factors, robust and biologically plausible microbial signatures emerge, offering more reliable candidates for clinical translation and future therapeutic development. The findings stress the need for standardized metadata collection, particularly stool form, diet, medications, and age, and show that strain-level resolution is critical for understanding gut–brain associations. This meta-analysis provides one of the clearest frameworks to date for interpreting ASD-microbiome research and highlights the heterogeneity inherent to ASD, supporting a shift toward phenotype-specific microbiome investigations.

Citation

West KA, Yin X, Rutherford EM, et al. Multi-angle meta-analysis of the gut microbiome in Autism Spectrum Disorder: a step toward understanding patient subgroups. Scientific Reports. 2022;12:17034. doi:10.1038/s41598-022-21327-9